BACKGROUND Production and health traits are central in cattle breeding. Advances in next-generation sequencing technologies and genotype imputation have increased the resolution of gene mapping based on genome-wide association studies (GWAS). Thus, numerous candidate genes that affect milk yield, milk composition, and mastitis resistance in dairy cattle are reported in the literature. Effect-bearing variants often affect multiple traits. Because the detection of overlapping quantitative trait loci (QTL) regions from single-trait GWAS is too inaccurate and subjective, multi-trait analysis is a better approach to detect pleiotropic effects of variants in candidate genes. However, large sample sizes are required to achieve sufficient power. Multi-trait meta-analysis is one approach to deal with this problem. Thus, we performed two multi-trait meta-analyses, one for three milk production traits (milk yield, protein yield and fat yield), and one for milk yield and mastitis resistance. RESULTS For highly correlated traits, the power to detect pleiotropy was increased by multi-trait meta-analysis compared with the subjective assessment of overlapping of single-trait QTL confidence intervals. Pleiotropic effects of lead single nucleotide polymorphisms (SNPs) that were detected from the multi-trait meta-analysis were confirmed by bivariate association analysis. The previously reported pleiotropic effects of variants within the DGAT1 and MGST1 genes on three milk production traits, and pleiotropic effects of variants in GHR on milk yield and fat yield were confirmed. Furthermore, our results suggested that variants in KCTD16, KCNK18 and ENSBTAG00000023629 had pleiotropic effects on milk production traits. For milk yield and mastitis resistance, we identified possible pleiotropic effects of variants in two genes, GC and DGAT1. CONCLUSIONS Multi-trait meta-analysis improves our ability to detect pleiotropic interactions between milk production traits and identifies variants with pleiotropic effects on milk production traits and mastitis resistance. In particular, this should contribute to better understand the biological mechanisms that underlie the unfavorable genetic correlation between milk yield and mastitis.

中文翻译：

---

在北欧荷斯坦奶牛中，从产奶性状和乳腺炎抗性之间的相关QTL区分多效性。

背景技术生产和健康特征是牛育种的中心。基于全基因组关联研究（GWAS），下一代测序技术和基因型估算的进步提高了基因定位的分辨率。因此，文献中报道了许多影响奶牛产奶量，牛奶成分和乳腺炎抗性的候选基因。带有效果的变异通常会影响多个特征。由于从单性状GWAS中检测重叠的定量性状基因座（QTL）区域太不准确和主观，因此多性状分析是检测候选基因变异的多效性的更好方法。但是，需要大量的样本才能获得足够的功效。多特征荟萃分析是解决此问题的一种方法。从而，我们进行了两项多特征荟萃分析，一项针对三种牛奶生产性状（牛奶产量，蛋白质产量和脂肪产量），一项针对牛奶产量和乳腺炎抵抗力。结果对于高度相关的性状，与单性状QTL置信区间重叠的主观评估相比，通过多性状荟萃分析提高了检测多效性的能力。通过多变量荟萃分析确定的铅单核苷酸多态性（SNPs）的多效性效应通过双变量关联分析得到了证实。先前报道了DGAT1和MGST1基因内变体对三种牛奶生产性状的多效性作用，并证实了GHR中变体对牛奶产量和脂肪产量的多效性作用。此外，我们的结果表明，KCTD16中的变体 KCNK18和ENSBTAG00000023629对牛奶生产性状具有多效作用。对于牛奶产量和乳腺炎抵抗力，我们确定了两个基因GC和DGAT1的变体可能具有多效性。结论多特征荟萃分析提高了我们检测乳汁生产性状之间多效性相互作用的能力，并鉴定了对乳汁生产性状和乳腺炎耐药性具有多效性影响的变体。特别是，这应有助于更好地理解牛奶产量与乳腺炎之间不利的遗传相关性的生物学机制。结论多特征荟萃分析提高了我们检测乳汁生产性状之间多效性相互作用的能力，并鉴定了对乳汁生产性状和乳腺炎耐药性具有多效性影响的变体。特别是，这应有助于更好地理解造成牛奶产量与乳腺炎之间不利的遗传相关性的生物学机制。结论多特征荟萃分析提高了我们检测乳汁生产性状之间多效性相互作用的能力，并鉴定了对乳汁生产性状和乳腺炎耐药性具有多效性影响的变体。特别是，这应有助于更好地理解造成牛奶产量与乳腺炎之间不利的遗传相关性的生物学机制。